I'd rather use a more proper bandpass filter than bp~, which does note have a nice symmetric response. You could use [biquad~] instead, but the issue is that you'd need to generate coefficients for it. I provide the [else/bicoeff] object in else that calculates biquad coefficients and I just uploaded an update in my tutorial that shows how to do that by hand in a pd patch, see: https://github.com/porres/Live-Electronic-Music-Tutorial/blob/master/Tutoria...
Em seg, 24 de jun de 2019 às 12:04, William Huston williamahuston@gmail.com escreveu:
I have one I built, it's basically 3x bp~ stages in parallel, with each stage being a double bp~ in series.
Get it here: https://github.com/WilliamAHuston/BHPDtoolkit
It's called: abstractions/BHFormantx3~.pd also abstractions/FormantControl.pd
I have a fake resonance built in, basically an osc~ at the cutoff freq of the filter which you can blend it.
It's not pretty, I built this as a PD novice. But it works well.
(You may want to grep through the patches folder to find some examples where I use this.)
One use I found for it, which is very cool, is to create a source which is rich in overtones, saw or square or a gritty granular snippet, send it to the filter.
Tune the 3 bp~'s to a chord, and use a midi controller to set the root of the chord, to play melodies *on the filter*.
HTH, BH
On Mon, Jun 24, 2019, 1:50 AM Ingo ingo@miamiwave.com wrote:
OK! I just googled "Formant Table" and got lots of results with the frequencies of each formant. I guess by using 3 bandpass filters I should be able to build one as an abstraction.
Ingo
Can anyone point me to a formant filter external that is working and not
too heavy on the CPU?
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list